Data · IPEDS + Scorecard + BEA
Retention is at a twelve-year high. What is behind it
First-year retention at four-year institutions reached 81.9% in fall 2024. Between a quarter and 36% of the rise comes from where students enrolled, the median institution gained one point, and the gap by Pell share grew.

Our overview of the 2024–25 IPEDS collection reported that first-year retention at four-year institutions rose from 79.5% in fall 2013 to 81.9% in fall 2024. That is the highest rate in the twelve years we have. This post takes the number apart: by year, by type of institution, by who enrolls, and by what moves with it.
IPEDS defines the retention rate as the share of first-time students who started the previous fall and are enrolled again, or have completed, by the current fall. At four-year institutions the cohort is students seeking a bachelor’s degree. Institutions report it in the Fall Enrollment survey, separately for full-time and part-time students.
Method: a year label is the fall students returned. “Fall 2024” is the 2024–25 IPEDS collection and covers students who started in fall 2023. National rates are weighted by the size of each institution’s full-time entering cohort, so they describe the typical student. Medians describe the typical institution. Sector and level are as each institution reported them that year.
The national rate, year by year
The four-year rate rose slowly from 2013 to 2020. It fell to 80.6% for students who started in fall 2020, about its 2014 level, and has risen each year since. Public four-year institutions are at 82.4%, the same as in fall 2020. Private nonprofits are at 81.8%, the same as in fall 2023.
All four-yearPublic four-yearPrivate nonprofit four-yearPublic two-year
Source: IPEDS Fall Enrollment, retention component, 2013–14 to 2024–25 collections. All institutions reporting a full-time rate, weighted by full-time entering cohort.
Public two-year colleges rose further, from 59.5% to 64.3%, but their high was fall 2023, at 64.5%. Four-year for-profits rose from 52.6% to 60.9%, on a cohort that fell from about 33,000 students to 23,000.
Between a quarter and 36% of the rise is where students enrolled
A national rate can rise because institutions retain more students, or because more students start at institutions that already retain most of them. Both happened.
Doctoral universities with very high research activity (R1) retain 89.6% of their first-year students. Their share of all full-time first-year students at four-year institutions grew from 33.7% to 38.3%. Meanwhile, 545 four-year institutions that reported a retention rate in 2013 did not report one in 2024. Their 2013 rate was 62.6%.
If each Carnegie class had kept its 2013 share of students, the 2024 rate would be 81.3% rather than 81.9%. Weighted the other way, the 2013 rates on 2024 shares give 80.4%. So the shift in where students enrolled accounts for 0.6 to 0.9 of the 2.4-point rise, or 26% to 36%. Among the 1,747 four-year institutions that reported in all twelve years, the rate rose from 80.1% to 82.1%.
By type of institution
| Carnegie class (latest) | Institutions, 2024 | Fall 2013 | Fall 2024 | Change |
|---|---|---|---|---|
| Doctoral: very high research (R1) | 145 | 87.8% | 89.6% | +1.8 |
| Doctoral: high research (R2) | 129 | 79.1% | 80.7% | +1.6 |
| Doctoral/professional | 173 | 74.9% | 76.2% | +1.3 |
| Master’s | 610 | 74.5% | 75.4% | +0.9 |
| Baccalaureate: arts and sciences | 215 | 84.7% | 84.5% | −0.2 |
| Baccalaureate: diverse fields | 263 | 67.1% | 69.8% | +2.7 |
| Baccalaureate/associate’s | 83 | 58.6% | 58.7% | +0.1 |
| Special focus four-year | 304 | 79.4% | 79.7% | +0.3 |
The gains are largest at research universities and baccalaureate colleges with diverse fields. Arts and sciences colleges started high and stayed flat. By size, institutions with 20,000 or more students retain 87.0%, up from 85.4%. Those under 1,000 retain 67.5%, up from 65.1%.
Selectivity sorts retention sharply. We grouped four-year institutions by their 2023 admission rate. Those admitting under a quarter of applicants retained 96.0% in 2023. Those admitting 90% or more retained 74.9%. Every admission-rate band rose by between 0.6 and 1.4 points over the decade, so the gap between them hardly moved.
The typical institution moved less than the national rate
The median four-year institution retained 72% of its full-time first-year students in fall 2013 and 76% in fall 2024. The 10th percentile rose from 44% to 53%, and the 90th from 91% to 94%. Small cohorts account for much of the spread at the bottom: among institutions with entering cohorts of 50 or more, the 10th percentile went from 54% to 59%.
Of the 1,747 institutions that reported every year, 886 retained a larger share in 2024 than in 2013 and 723 a smaller one. The median change was a gain of one point, but only 354 institutions (20%) were within one point of their 2013 rate. Only 317 were at or tied with their own twelve-year high in 2024. A national high is not a high for most institutions.
Small cohorts swing. At institutions with fewer than 100 full-time entering students, the median year-to-year change was 10 points. With 2,000 or more students, it was 1 point. A one-year change at a small college is mostly noise.
Pell share and who enrolls
IPEDS does not report retention by Pell status, race or income. The table built for a gap by race in our warehouse is empty for that reason. What we can compare is institutions: those that enroll more Pell recipients against those that enroll fewer.
Source: IPEDS Fall Enrollment, 2023–24 collection (retention); College Scorecard share of undergraduates receiving Pell grants, as loaded for 2023–24. 2,063 four-year institutions in four equal groups.
The gap widened over the decade. We took the 1,815 four-year institutions with a Pell share that reported a rate in both 2013 and 2023 and placed each in a band by its 2023 Pell share. The 535 institutions under 25% Pell rose from 86.4% to 88.1%. The 247 institutions at 55% or more stayed flat, at 71.9% and 72.0%. The difference grew from 14.5 to 16.1 points.
By race and ethnicity the top and bottom differ in the same direction, though not step by step. Four-year institutions where 46% or more of students are Black, Hispanic, American Indian, Pacific Islander or of two or more races retained 74.2% in fall 2024. Institutions under 18% retained 83.9%, and those between 18% and 28% retained 85.7%. Over the same period that share of four-year enrollment rose from 28.1% to 36.0%. Retention rose while the student body changed. These are figures about institutions, not about students: a lower rate at a high-Pell institution does not tell us how its Pell students fare compared with others.
Part-time students and two-year colleges
Part-time first-year students at four-year institutions are retained at 49.2%, up from 44.2%. At public two-year colleges the part-time rate is 44.5%, up from 43.1%, after a low of 40.8% in fall 2020. That is about 20 points below the full-time rate at the same colleges.
Public two-year colleges also got smaller. Their full-time entering cohort fell from 698,000 students who started in fall 2012 to 457,000 who started in fall 2023 (−34.6%). The part-time cohort fell from 464,000 to 289,000 (−37.8%). The full-time cohort at four-year institutions grew 8.7% over the same years. Among 814 public two-year colleges that reported every year, full-time retention rose from 59.8% to 64.4%. The colleges did retain more of their entrants. Whether the students who stopped enrolling would have persisted at lower rates, IPEDS cannot say.
The local economy
We matched four-year institutions to their county’s per-capita personal income for 2023 from the Bureau of Economic Analysis. We ranked institutions by their county’s income. The quarter with the lowest county income, under $55,109 per person, retained 79.0% in fall 2024. The quarter at $77,192 or more retained 85.3%. The middle two quarters were about equal, at 81.8% and 81.2%. County income travels with selectivity and wealth, so this is a description, not a local effect. 117 of 2,073 institutions have no matched county income and are left out: Puerto Rico and the territories, Connecticut’s new planning regions, and some Virginia independent cities.
What moves with retention
Our relationships table tests every pair of measures across institutions in the 2023–24 survey year, controlling for undergraduate enrollment. It also tests year-over-year changes within institutions. A row means the pair passed a significance and size test. It never means one causes the other.
| Measure | Relationship | Institutions compared | Coefficient | n | Tier |
|---|---|---|---|---|---|
| Average SAT score | moves with | All | 0.74 | 1,071 | direction |
| Part-time retention rate | moves with | All | 0.41 | 2,911 | direction |
| Share of students exclusively online | moves against | Public four-year | −0.39 | 670 | direction |
| Pell share | moves against | Associate’s: high transfer, high traditional | −0.51 | 103 | direction |
| Share of students who are Black | moves against | Master’s, larger programs | −0.38 | 307 | direction |
| Student services spending per FTE | moves with | Baccalaureate/associate’s | 0.44 | 48 | direction |
| Academic support spending per FTE | moves with | Baccalaureate/associate’s | 0.47 | 48 | direction |
| Total spending per FTE (change, 2014–2023) | changes with | Arts and sciences colleges | −0.25 | 2,166 (221 inst.) | exists |
| County per-capita income | moves with | Associate’s: high transfer, mixed traditional | 0.30 | 89 | exists |
Entering preparation leads the list. Institutions with higher average SAT scores tend to retain more first-year students, at 0.74 nationally (n = 1,071) and 0.90 among R1 universities (n = 127), both direction tier, 2023–24. Among public four-year institutions, those with more exclusively online students tend to retain fewer (moves against, −0.39, n = 670, direction tier, 2023–24).
Spending shows fewer relationships with retention than entering preparation does. Instruction spending per FTE has no relationship found with full-time retention in any group of institutions. That means it did not pass the test, or it was not among the twelve strongest neighbors of either measure. Student services and academic support spending move with retention only among 48 baccalaureate/associate’s colleges (0.44 and 0.47, n = 48, direction tier, 2023–24).
The one change row runs the other way. At arts and sciences colleges, years when total spending per FTE rose tended to be years when retention fell (changes with, −0.25, 2,166 institution-years, 221 colleges, 2014–2023, exists tier). One possible reason is arithmetic: when students leave, the same budget is divided by fewer FTE. The table cannot say whether that is what happens. In the same group, undergraduate enrollment changes with retention at 0.27 (2,172 institution-years, 221 colleges, 2014–2023, exists tier).
Tiers: “direction” means Pearson and rank coefficients agree in sign and the same relationship passed in at least three groups; “exists” means it passed the test only. Spending figures are for fiscal 2023, the year before the 2023–24 survey year. The two rows for baccalaureate/associate’s colleges use the College Scorecard’s delivery of the full-time retention rate; every other row uses the IPEDS rate. A relationship across institutions in one year is not a trend, and it says nothing about what would happen if one institution changed its spending.
What the data cannot tell
- Retention for Pell recipients, first-generation students or any racial group at a given institution. IPEDS collects only the overall full-time and part-time rates.
- Whether a student who left transferred and kept going. A student who transfers counts as not retained.
- Why the rate rose at a given institution. Advising, aid, admissions policy and the mix of majors all leave the same trace in this number.
- Whether students who did not enroll after 2020, especially at two-year colleges, would have been retained at the same rate as those who did.
What this means for planning
- Compare your retention with institutions of the same Carnegie class, admission rate and Pell share. More than half of the students behind the national 81.9% (55.5%) are at R1 and R2 universities.
- Read a one-year change against your cohort size. Below 100 entering students, a 10-point swing is ordinary.
- Report part-time retention next to full-time. At two-year colleges the gap is 20 points.
- Test a spending plan against your own history before relying on it. In our relationships table, instruction spending per FTE shows no relationship found with full-time retention in any group of institutions, 2023–24.
- Track your own Pell and first-generation retention internally. The federal data will not show your gaps.
Every figure here comes from public data: IPEDS, the College Scorecard and the Bureau of Economic Analysis. In Academia Superiora you can run the same comparisons for your own peer group, free.
